combat decision making process at the tactical level

7/28/2019 Combat Decision Making Process at the Tactical Level

1/156

DEFINING A COMBAT DECISION-MAKING PROCESS AT THE TACTICALLEVEL OF WAR AND OPERATIONS OTHER THAN WAR

A thesis presented to the Faculty of the U.S. ArmyCommand and General Staff College in partial

fulfillment of the requirements for thedegree

MASTER OF MILITARY ART AND SCIENCE

B. DON FARRIS 11, MAJ, USAB.S., United States Military Academy,

West Point, New York, 1983

Fort Leavenworth, Kansas1995

Approved for public release; distribution is unlimited.


2/156

8WO NW(NOW

0SEaAN.

8OOp


3/156

MASTER OF MILITARY ART AND SCIENCETHESIS APPROVAL PAGE

Name of Candidate: MAJOR B. DON FARRIS I1Thesis Title: Defining a Combat Decision-Making Process at the Tactical

Level of War and Operations Other Than War

Thesis

Member

Member

Chairman

Accepted this 2nd day of June 1995 by:

, Director, Graduate DegreePhilip J. Brookes, Ph.D. Programs

The opinions and conclusions expressed herein are those of the studentauthor and do not necessarily represent the views of the U.S. ArmyCommand and General Staff College or any other governmental agency.(References to this study should include the foregoing statement.)


4/156

ABSTRACT

DEFINING A COMBAT DECISION-MAKING PROCESS AT THE TACTICAL LEVEL OF WARAND OPERATIONS OTHER THAN WAR by MAJ B. Don Farris 11, USA, 147pages.

This study challenges the deliberate decision making process (DDMP) asineffective for use by conunanders in time-constrained environments. FM101-5 (final draft), Command and Control for Commanders and Staff, 1995,currently prescribes DDMP as the only accepted process at the tacticallevel for both war and operations other than war. The researchersuggests a second, quicker and more flexible doctrinal methodology tosupplement DDMP when planning time is limited.DDMP relies on a traditional, analytical approach of generating andconcurrently evaluating options allowing commanders to make an "optimal"decision. This study explores the existing theory and relevance of . -"satisficing." When planning time is limited, the initial course ofaction developed by the commander may prove more effective than onearrived at after the careful consideration of options. This theoreticalapproach provides the framework for a proposed combat decision makingmodel. Using the doctrinal characteristics of an effective decisionmaking process, both DDMP and the proposed model are compared. Thestudy concludes that the proposed combat decision-making process shouldreplace DDMP in fluid, time-constrained environments.

iii


5/156

ACKNOWLEDGMENTS

To my wife, Tara, who's love, patience and encouragement madethis work entirely possible.

To my MMAS Committee and the MMAS Staff, who answered a host ofquestions, kept me on azimuth, and who's advice and counsel provedinvaluable throughout each phase of the process.

To the Observer/Controllers at the Joint Readiness TrainingCenter, Fort Polk, Louisiana, who's tireless efforts continue to makethe Army better prepared for the next fight.


6/156

TABLE OF CONTENTS

APPROVALPAGE . . . . . . . . . . . . . . . . . . . . . . . . . . . iiABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiiACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . ivLIST OF ILLUSTRATIONS . . . . . . . . . . . . . . . . . . . . . . . viCHAPTER

1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . 12. REVIEW OF LITERATURE . . . . . . . . . . . . . . . . . . . . 123. A PROPOSED COMBAT DECISION MAKING PROCESS . . . . . . . . . . 464. ANALYSIS . . . . . . . . . . . . . . . . . . . . . . . . . . 685 . CONCLUSIONS AND RECOMMENDATIONS . . . . . . . . . . . . . . . 102

APPENDIX. EXAMPLE OF BRIGADE STAFF BATTLE DRILLS . . . . . . . . . 110BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . 144INITIAL DISTRIBUTION LIST . . . . . . . . . . . . . . . . . . . . . . 146


7/156

LIST OF ILLUSTRATIONS

Figure 1. Annotated Steps of the Deliberate DecisionMaking Process . . . . . . . . . . . . . . . . . . . 2Figure 2. Annotated Steps of the Proposed CombatDecision Making Model . . . . . . . . . . . . . . . 47. . . . . . .igure 3. The Deliberate Decision Making Procedure 49Figure 4. Subordinate Commanders' Input IntegratedWithin the Proposed Combat Decision MakingModel . . . . . . . . . . . . . . . . . . . . . . . 51Figure 5. Commander's Course of Action IntegratedWithin the Proposed Combat Decision MakingModel . . . . . . . . . . . . . . . . . . . . . . . 54Figure 6. War Gaming and Synchronization Integrated

Within the Proposed Combat Decision MakingModel . . . . . . . . . . . . . . . . . . . . . . . 56Figure 7. Targeting Integration Step Integrated Within . . . . .he Proposed Combat Decision Making Model 59Figure 8. Commander's Confirmation Integzated Within . . . . .he Proposed Combat Decision Making Model 61Figure 9. Execution and Continuous Assessment IntegratedWithin the Proposed Combat Decision MakingModel . . . . . . . . . . . . . . . . . . . . . . . 63Figure 10. Example of the Deliberate Decision Making

Process . . . . . . . . . . . . . . . . . . . . . . 71Figure 11. Example of the Deliberate Decision MakingProcess . . . . . . . . . . . . . . . . . . . . . . 75Figure 12. Example of the Deliberate Decision MakingProcess . . . . . . . . . . . . . . . . . . . . . . 82Figure 13. Example of the Proposed Combat DecisionMakingModel . . . . . . . . . . . . . . . . . . . . 83Figure 14. Assigning Variables to the Deliberate Decision. . . . . . . . . . . . . . . . . . .aking Process 84Figure 15. Assigning Variables to the Proposed CombatDecision Making Model . . . . . . . . . . . . . . . 84. . . . . . .igure 16. The Proposed Combat Decision Making Model 98Figure 17. React to Enemy Ground Contact . . . . . . . . . . . . . 113. . . . . . .igure 18. React to Enemy Indirect Fire (Q36 Drill) 117


8/156

F i g u r e 1 9. R e ac t t o Enemy M i l i t a r y I n t e l l i g e n c eI n t e r c e p t . . . . . . . . . . . . . . . . . . . . . 2 1F i g u r e 2 0 . R ea c t t o Enemy A i r D efense . . . . . . . . . . . . . . 25F i g u r e 2 1 . R e ac t t o Enemy L o g i s t i c s S i t e . . . . . . . . . . . . . 29F i g u r e 2 2 . R ea ct t o A i r T h r e a t . . . . . . . . . . . . . . . . . . 33F i g u r e 23 . R e a c t t o E n e m y M i n e f i e l d . . . . . . . . . . . . . . . 37F i g u r e 2 4 . R e ac t t o C he mi ca l A t t a c k / T h r e a t . . . . . . . . . . . . 40

v i i


9/156

CHAPTER ONEINTRODUCTION

"A Poor Plan Executed On Time Is Far Better ThanThe Perfect Plan Executed Too Late!"

Old Military Maxim

BackgroundField Manual 101-5, Command and Control for Commanders and

Staff, is the principal doctrinal tool used by the U.S. Army to teach,coach and train commanders and staff on the process of how to makeeffective decisions in combat. Since 1932, the Army has published nineversions of this manua1.l With each subsequent revision, the UnitedStates Army's decision-making process continues to evolve. Today, atenth version is currently under review by the Army's leadership indraft FM 101-5.2

In chapter four of draft FM 101-5, the doctrine describes onlyone decision-making process--the deliberate decision-making process(DDMP) 3 This single dec~sion-making rocess is intended forapplication in any environment. In those situations where time islimited, a commander may modify or abbreviate the process as he deemsappropriate.4 Any abbreviation or modification of the process, however,must not arbitrarily skip any of the nine annotated steps.5 These nineprinciple steps of the deliberate decision-making process as currentlydefined in draft FM 101-5 are shown in Figure 1 on the next page.6


10/156

DELIBERATE DECISION MAKING PROCESS

Fig. 1. Annotated Steps of the Deliberate Decision-Making Processextracted from ST 101-5, Command and Staff Decision Processes, pg 1-2-5.

The recent advent of the Army's combat training centers (CTCs)and the Battle Command Training Program (BCTP) has placed great emphasison training commanders and staffs to apply the deliberate decision-making process. The CTCs focus on planning, preparation, and executionat the brigade level and below, while BCTP focuses on planning andexecution at the division and corps level using a computer simulationmodel. In each case, the participating commanders and their staffs are


11/156

exposed to the stressful and realistic conditions of decision-making insimulated combat conditions.

In the last several years numerous observations from the field,interviews, and empirical research have attempted to assess the Army'soverall progress. Evidence exists to indicate that commanders and staffhave difficulty applying the deliberate decision-making process both incombat and during training at the CTCs and BCTP. Whether this observeddifficulty is due to training and experience of the participants or aflawed process continues to be hotly debated. Regardless of which sideof the debate that' leaders and researchers fall into, the vast majoritydo acknowledge that the "time available" to complete the process alwaysremains crucial.

Applying any decision-making process is especially problematicin crisis, fluid or time-constrained environments. Draft E'M 101-5 evenaccepts that there may be situations where it is impossible to use thecurrent DDMP.' Again, when these situations occur, the commander isexpected to abbreviate or accelerate the planning process as he deemsnecessary. Yet, beyond acknowledging that these situations will occur,doctrine does not offer any "detailed" discussion of how to effectivelyabbreviate the process, nor does it define any suitable alternatives.

Crisis, fluid, and time-constrained environments would seem tobe realistic descriptions of the nature of combat. If the readeraccepts this observation, either through personal experience or theexperience of others, then the Army's doctrine may be lacking. It maynot be advisable to train commanders and staffs on a single decision-making process that doctrine accepts in advance will not hold up in many


12/156

combat situations. If this assessment of the doctrine proves accurate,then a search for alternatives is both necessary and prudent.

The current field of research in military decision-makingpurports that two distinctly different cognitive strategies exist--optimizing and ~atisficin~.~n "optimizing" strategy is considered themore traditional analytical approach and characterizes the current DDMP.All availableinformation is used to generate conceptual possibilitiesthat will solve a particular tactical problem. The decision makerconcurrently evaluates all the generated options and selects what hebelieves to be the best solution. This "best" solution is consideredthe "optimal" choice under the given c~nditions.~

A "satisficing" strategy challenges the traditional analyticalapproach. Satisficing advocates that experienced decision makers do notgenerate or conceptualize numerous options and then evaluate themconcurrently; especially in stressful, time-constrained environments.Satisficing theory states that a decision maker considersall availableinformation, looks for certain informational queues within the givenenvironment, and relies on his experience to generate a single optionthat satisfies the conditions. This first developed option is arrivedat much quicker than an optimal approach that must wait for all optionsto be developed and evaluated before making a decision.1

If the current DDMP with its "optimizing" approach provesinadequate in time-constrained environments, then a "satisficing"approach may prove a suitable alternative. This thesis will explorethis premise and attempt to present a second, more rapid "combat


13/156

decision-making process" that provides greater flexibility forcommanders.

Problem Statement and Research QuestionsThe 1995 draft of EM 101-5, Command and Control for Commanders

and Staff, describes a single decision-making process. The proposeddraft doctrine also acknowledges that DDMP may become difficult orimpossible to use in some time-constrained situations. If combat isoften characterized by a lack of sufficient planning time, then theArmy's doctrine may be inadequate if it does not address a sufficientlydetailed alternative.

The primary research question to answer is this: Would anadditional "combat decision-making process" significantly improve theability of commanders and staffs to make decisions in time-constrainedenvironments? To further amplify the primary research question, thefollowing secondary research questions must also be addressed:

1. Is the current deliberate decision-making process tooinflexible to meet the needs of commanders and staffs in time-constrained environments?

2. Does the proposed doctrine adequately address in sufficientdetail how to abbreviate the current deliberate process to compensatewhen planning time is limited?

3. If the current deliberate process is too inflexible, whatshould substitute as an effective combat decision-making model, andunder what conditions should it apply?


14/156

Assumptions1. For the purpose of comparative analysis, the best test of

any decision-making process is under actual combat conditions involving

numerous commanders and staffs. Obviously, this type of test of anyproposed combat decision model is unfeasible.

2. The United States Army will continue to select commander'swho are generally more experiencedat decision-making than their staffsor subordinate commanders.

Definition of Key TermsThe definitions used in this thesis are extracted from ST 101-5-

1 (draft), Operational Terms and Symbols, and are as follows:llBranch. A contingency plan or course of action (an option built

into the basic plan/course of action) for changing the mission,disposition, orientation, or direction of movement of the force to aidsuccess of the operation based on anticipated events, opportunities, ordisruptions caused by enemy actions and reactions as determined duringthe war gaming process.

End State. A set of conditions which, when achieved, attain theaims set for the campaign or operation.

Fragmentary Order (FRAGO). An abbreviated form of an operationsorder (OPORD) that eliminates the need for restating informationcontained in a basic operation order. It may be issued in sections.

Operations Order (OPORD). A directive issued by a commander tosubordinate commanders for effecting the coordinated execution of anoperation.


15/156

Operations Other Than War (OOTW) Military activities duringpeacetime and conflict that do not necessarily involve armed clashesbetween two organized forces.

Sequel. Major operations which follow the current majoroperation. Plans for these are based on the possible outcomes -victory, stalemate, or defeat - associated with the current operation.scope

The proposed combat decision-making process presented in thisthesis is primarily designed for United States Army commanders andstaffs operating at the tactical level of war and operations other thanwar--namely, corps and below. The proposed combat decision-making modeldoes not attempt to address the operational or strategic level.

The commander's intuition, ability, and experience are verygermane to the research question. No single process, no matter howeffective, can ever hope to overcome the failure of a commander tocorrectly identify what is decisive and quantify it in terms of anattainable end state. The commander's ability to correctly identify the.problem or desired end state is perhaps more "art" than "science." Assuch, the failure of a commander to correctly visualize the desiredoutcome should not then become an indictment on any combat decision-making process as either unsound or ineffective. This thesis will notattempt to address the commander's intuitive ability--only the process.

Additionally, perceived difficulties with the application ofDDMPmay indicate more than a flawed process. How the Army trainscommanders and staffs on DDMP may present additional concerns.


16/156

Increased deployments and high personnel turnover during the Army's drawdown can certainly have an impact and limit training opportunities. Thecurrent field of research may not clearly distinguish where problemswith the process end and the training problems begin. Again, thisthesis will not attempt to address the Army's training environrnent--onlythe process.

ImportanceThe ability to make quicker and better decisions in combat has

significant relevance to accomplishing future missions with fewerAmerican lives lost. If the draft doctrine is incomplete, then effortsto supplement the Amy's decision-making process in combat may proveextremely beneficial. The United States' recent involvement in Somaliaand Haiti is a perfect example of how both the environment and end statecan change with very little time to react or adjust the initial plan.Crisis, fluid, and time-constrained environments should be expected.This, in the author's opinion, is the most likely scenario thatcommanders and staffs will continue to find themselves in during thenext several years.

MethodologyThe initial stage of this research project required finding

sufficient evidence to support the author's opinion that a problemexists with the current deliberate decision-making model. The review ofliterature presented in chapter two focuses to the fullest extent onexisting observations and empirical studies conducted by both militaryand civilian researchers. The observations and research are organized

8


17/156

into six separate categories and were collected in laboratory,classroom, garrison, and field settings using a variety of methods,structured interviews, and controlled experiments. Within the scope ofcognitive human performance research, the inability to predict and limitthe number of variables affecting a given outcome will remain elusive.As such, efforts to affix one or more specific factors to a documentedreason for failure is subject to interpretation.

Additionally, there is a great deal of research and much iswritten about decision theory, but as it applies to military operations

a relatively small amount is documented. An overwhelming majority ofthe research is conducted by the Army Research Institute at the behestof the Army's leadership. This may cause some to infer a slight bias inthe available research. A m y Research Institute (ARI) researchers,however, do tend to target perceived problem areas previously identifiedb y leaders and the force as a whole. Despite the limited field ofstudy, a sufficient quantity and quality of empirical research andobservations do exist to provide evidence of a number of deficiencies inthe current deliberate model.

Using the lessons of experience provided by the CTCs andbuilding on the application of Recognition-Primed Decision Making,12discussed later in this thesis, the second stage of the research is tobuild a proposed combat decision-making model. Each step of theproposed model is presented along with a discussion of its importance,anticipated application, and how it differs significantly from thedeliberate process. It is intended that the proposed combat decision-making model will provide sufficient detail for use by commanders and

9


18/156

staffs in lieu of the deliberate model. Any elements of the proposedcombat decision model that remain unchanged from current doctrine arenot reiterated for the sake of brevity.

The third stage of the research is a comparative analysisbetween the proposed combat model and the deliberate process. Theanalysis isorganized into five categories: (1) a discussion of the twoexisting theoretical approaches to military decision-making; to includethe Recognition-Primed Decision (RPD) methodology; ( 2 ) a comparativetime analysis of the two models; (3) allowing for subordinatecommanders' input; (4) the value of formalizing the targeting processwithin the decision model; and (5) a discussion on how the proposedcombat decision model places greater emphasis on keeping the processcontinuous and cyclic--the importance of the execution step in combatdecision-making.


19/156

Endnotesl ~ e x. Michel, Historical Development of the Estimate of the

Situation (Arlington, VA: U.S. Army Research Institute for theBehavioral and Social Sciences, 1990), 1.

2 ~ . ~ .rmy, FM 101-5 (draft), Command and Control for Commandersand Staff (Washington, DC: Department of the ~rmy, 995).

%.s. Army, ST 101-5, Command and Staff Decision Processes (FortLeavenworth, KS: Command and General Staff College, 19931, 1-2-5.

8~ary . Klein and Roberta Calderwood, Investigations ofNaturalistic Decision Making and the Recognition-Primed Decision Model,(Arlington, VA: U.S. Army Research Institute for the Behavioral andSocial Sciences, 1990).

llU.S. Army, ST 101-5-1 (final draft), Operational Terms andSymbols (Washington, DC: Department of the Army, 1995), 1-248.

12~lein nd Calderwood.


20/156

CHAPTER TWOREVIEW OF LITERATURE

An overview of the existing research oK tactical planningprocess was recently completed in 1993 by Jon Fallensen of the ArmyResearch 1nstitute.l Using a modified version of his original format,this review of the literature on the tactical planning process isorganized into seven categories: (1) historical perspective, ( 2 )failure to follow procedures, (3) management of the process, ( 4 )inflexible procedures, ( 5 ) excessive time demands, (6) formulation ofalternatives, and (7)war gaming and synchronization.

Primary sources of the information available come fromobservations at the Combat Training Centers (CTCsl, the Battle CommandTraining Program (BCTP), the Center for Army Lessons Learned (CALL), theArmy Research Institute (ARI), after action surveys from DESERT STORM,the Army Battle Command Battle Lab, and a variety of independent studiesfrom Horizons Technology, Institute for Defense Analyses (IDA), and theRand Corporation. It should be noted that observations from the CTCswere not comprehensive, but assembled by CALL observers during twoseparate time periods. The first collection of observations weredeveloped from the Army Lessons Learned Management Information System(ALLMIS) during 1986-1989. The second collection of observations werecompiled by CALLobservers during 1991-1992.


21/156

Historical PerspectiveMichel's (1990) research on the historical development of the

estimate process points to the Prussian's as the first to develop a

formal methodology to aid commanders in making decisions on thebattlefield. This first attempt to provide a systematic and analyticalapproach to military problem solving continued to evolve in Europe untilthe early 1900's when it was first introduced at the Infantry andCavalry School at Fort Leavenworth. By 1910, the United States Armyadopted the estimate of the situation as official doctrine. Then, in1940 the first EM 101-5, Staff Officer's Field Manual, the Staff andCombat Orders, was released using a five paragraph commander's estimateconsisting of mission, situation and courses of action, analysis,comparison, and decision. Over the past 50 years, E'M 101-5 has beenrevised a total of nine times, not including an unpublished draft in1977. As of this date, FM 101-5 is currently being reviewed for itstenth formal revision.

While the format for the estimate process has remainedessentially the same since 1940, the practical procedures for itsapplication has evolved considerably. Four specific areas surroundingthe evolution of this process are germane to the literature review: (1)a mental versus written process; (2)who conducts the estimate; (3) theinfluence of time on completing the process; and ( 4 ) generating optionsfor friendly courses of action.

Mental versus written process. The original estimate processwas initially considered a mental process. Michel (1990) states:

Mention of procedures for performing the estimate was limitedto considering the estimate steps as a "train of thought sequence"


22/156

in the 1932 version. In that first issue it was also mentionedthat an estimate may be a purely mental (as opposed to written)exercise. In the 1940 version, this recognition was expanded toconclude that a "mental process" was the most commontype ofestimate at division level and below. This conclusion has beencarried through every revision since then.3

While the original intent of the estimate process to serve as a "mentalroadmap" remains unchallenged, the preeminence of the estimate processas a product-oriented methodology began primarily in the last decadewith the publication of CGSC ST 100-9, The Command Estimate Process.This publication, until it was replaced in 1993 by CGSC ST 101-5,Command and Staff Decision Processes, served as the practical guide fortraining commanders and staffs throughout the Army.

While not official doctrine, CGSC ST 101-5 continues to serve asthe accepted step-by-step application of the deliberate decision-makingprocess. The inclusion of several written or drafted IPB (IntelligencePreparation of the Battlefield) products, the listing ofspecified/implied/essential tasks, a written "restated" mission, thecommander's written intent statement, COA sketches, COA statements,formal COA comparison, and the recent development of the synchronizationmatrix are examples of how the "written process" now overshadows theoriginal "mental process.

Who conducts the estimate. Michel's (1990) research furtherhighlights that the estimate process was originally designed for theexclusive use of commanders. Not until the 1950 version of EM 101-5would the estimate process expand to encompass the staff. Ten yearslater in the 1960 version, the "military decision-making process" wasfirst introduced and the staff began to retain primary responsibilityfor developing the overall estimate "for the commander." The

14


23/156

commander's own estimate was incorporated as a smgle step in theoverall estlmate process. Michel states:

This formal process places the commander's estimate within thebroader context of all the command and staff actions required todevelop and execute a course of action. . . . Thus, the delineationof this process helped make obvious the fact that the commanders'estimate is often an interactive undertaking in which his staffplays a very active part.4

Further steps were taken in the 1968 version to direct which members ofthe staff were responsible for each specific step in the estimateprocess. Michel records that this level of detail in the process wasdebated at the Command and General Staff College at great lengths:

The 1968 version made the first real effort to distinguish theresponsibilities of the commander, G2 and G3 in the generation ofpossible courses of action, a problem that had been worked on atthe Command and General Staff College for some time.5

Only the 1977 draft version differed from this specificbreakdown of responsibilities for each staff member. Michel's researchshows that this draft version, in sharp contrast to the earlier 1968version, placed much of the actual mechanics of the decision processsquarely back on the shoulders of the commander. The commanderdeveloped the courses of action and "mentally" war gamed each one. Thenecessity of making a decision within a time-constrained, combatenvironment was offered as the primary reason for keeping the decisioneffort with the commander. The focus of the estimate process with moredetailed analysis was shifted back to the staff, however, when timepermitted.

The influence of time. Throughout its development the impact oftime on the estimate process also emerged as an issue, with the greatestemphasis occurring shortly after the Korean and Vietnam wars. During

15


24/156

World War I1 the estimate process does not mention the impact of time,perhaps largely because during this era it remained primarily a mentalprocess. After the Korean War, however, the 1954 version of FM 101-5makes mention of the necessity when time is limited to "combine" steps.Michel states:

The post-Korean war 1954 release contains a unique proceduralstatement. In a paragraph entitled "Basic consideration" it issuggested that under certain circumstances an estimator mightcombine certain elements of the basic outline. This suggestion isnot repeated in subsequent releases.

Two revisions later and during the height of the Vietnam War the

1968 version contained a similar statement indicating the need forflexibility:

The 1968 manual also contained a caveat about using theestimate process that was somewhat like the one that appeared in1954, but did not go quite as far. The statement was made that theformat was not rigid. An estimator could go on to the next stepwithout completing the preceding one or make several smallexcursions through the process within his overall a~timate.~

The most notable mention of the effect of time on the estimateprocess occurs in the unpublished 1977 version. Less than five yearsafter the Vietnam War, and presumably with input shaped by combatexperience, the 1977 version took a radical departure from the existingdoctrine. Michel states:

The 1977 draft viewed the commander's estimate as a much moredynamic, subjective and hurried process than any version before orsince. . . . It viewed the structured steps in the commander'sestimate and in the military decision-making process as fullyapplicable only when time was available, but typically "speed isthe essence of the process in the fast-moving en~ironment."~

This version, which apparently ran counter to the accepted viewpoints onthe estimate process, was never adopted.


25/156

Generating options. The practice of developing multipleoptions, while fairly consistent in format, has certainly evolved in itsapplication with each subsequent revision since 1932. A detaileddescription of how to develop multiple courses of action was notintroduced until the 1940 version. This version provided the model thatwould be subsequently built upon in succeeding revisions, but providedtwo qualifying statements that would eventually disappear after the 1950revision:

One was that if only one practical line of action seemed open toyou, you could proceed directly to the decision. . . . The secondwas that, "as a general rule, not more than two or three own linesof action need to be carried forward for further anal~sis."~

The 1954 version further expanded the necessity of generatingmultiple options, and begrudgingly accepted-the idea that generating asingle course of action could suffice:

the commander in visualizing his possible courses of action . . . ,"eliminates from further consideration those . . . which areobviously inferior to the others being considered." Itcontradicted the 1940 version in stating that although in certaincircumstances there may be only one practical course of action, thefinal decision would not be made until that course of action istested . . . to determine its "ramifications.".. 10

The greatest emphasis on a detailed look at multiple optionsbegan in the 1968 version where the term "war gaming" was first appliedto the process. This version listed eight products that should resultfrom its detailed war gaming methodology. Most significant, however,were the instructions that the war gaming process would be repeated foreach course of action considered. This monumental leap in both processand written requirements did not discuss the nature of the environmentin which these deliberations would take place or the imediate


26/156

consideration of how time might influence the process. Thus, the 1968version began the debate that largely continues today, by espousing theneed for flexibility, yet advocating in the same document the need for adetailed, written wargaming of multiple options--an unprecedentedrequirement that certainly lengthened the process.

Michel further examines that since this 1968 version, there wasa "leveling off" of the detail required in the process until the 1986publication of CGSC ST 100-9. As mentioned earlier, this publicationprescribes a significant increase in the formal products and writtenrequirements necessary to complete the D D M P . ~ ~

Summary. Precisely why and what considerations weighed heavilyon modifications to each subsequent revision of the estimate process isunknown. The United States Army does not have a method of recording thedeliberations of each set of authors--only the finished doctrinalproduct. What does remain clear after nine revisions, and a tenth nowin consideration, are the trends that have evolved. What began largelyas a mental process solely oriented at commanders for achieving abattlefield solution has now evolved into a time intensive, staff drivenprocess. It is interesting to note that shortly after the Korean andVietnam Wars the greatest resurgence to make the commander principallyresponsible for the estimate process occurred. It is also important toobserve that during these same periods the necessity for flexibilityreceives its greatest emphasis. Yet, at the height of the Vietnam Warin the 1968 version, the decision-making process made its transitioninto a staff driven, product-oriented process that largely exists todayin draft EN 101-5.

18


27/156

Vietnam, it should be observed, was the only conflict inAmerican history to adopt a policy of rotating commanders every sixmonths at the tactical level. ~ i k e he Prussians over 200 years ago,perhaps the perception of inexperience created its own safety net--moreprocess. Just five years after Vietnam, and presumably after much soulsearching from their combat experience, the authors of the 1977 draftversion attempted to return the estimate process back to a commanderoriented, cognitive process. A process that relied on the strength ofan experienced commander and placed a premium on the quickest possible

decisions under combat conditions. It was never adopted.

Failure to Follow ProceduresObservations at the CTCs and ARI studies provrde evidence that

the current DDMP is not routinely followed. In each case it wasdifficult to capture the exact reasons why commanders and staffsdeviated from the prescribed doctrinal process. The dynamics of groupbehavior combined with an overwhelming variety of cognitive abilitiesundoubtedly preclude any researcher from isolating all the variables.One could argue that it is precisely this influence of so many variablesthat the need for greater flexibility in the process exists. The moreoften deviations from any process occur, the stronger the inference thatthe process may be inadequate.

Observations extracted from ALLMIS (1986-1989) by CALL analystsindicate that in 33 rotations at NTC, JRTC, and BCTP approximately one-third of the observations are directly attributed to a failure to followthe doctrinal planning process.12 Further observations by battlefield


28/156

operating system compiled by CALL (1991-921 at the CTCS illustratenumerous problems with correctly following the process:

The ability of task force staffs to foresee events on thebattlefield, through wargaming, is not understood or completed insufficient detail. . . . IPB is not conducted to standard becausethe 52 is usually the only staff officer actively involved in itsdevelopment. 3

Performance trends from BCTP, also complled by CALL (19921, furthershowed that 76 percent of the staffs did not develop viable plans orconduct parallel planning.14

Michel and Riedel (1988) investigated the effects of expertiseand cognitive style in tactical decision-making problems by ~reatingsimulated task of course of action development and evaluation at thedivision level. Using eight lieutenant colonels (experienced group) andeight CGSC students (novice group), each group was asked to developseparate concepts of operation for an offensive and a defensive mission.The purpose of this experiment was to investigate the differences inexpertise, cognitive styles, and use of information between the twogroups, and how each contributes to the decision-making process. Thefindings showed that neither test group felt confined by the doctrinalprocess:

Participants generally did not approach this problem with adefinite schema in mind for solving it. There was a high degree ofinterest in the task but also there was evidence of someexperimentation going on as suggested by participant statementslike, "I know I'm suppose to do it X way, but I want to try Y." . .. The lack of standard schemas in determining a course of actionmight therefore be more indicative of battlefield decision-makingthan had we found a "lock step" use of data.15

Fallensen, Carter, et al. (1992) conducted a controlledexperiment with fourteen pairs of experienced field-grade Army officers


29/156

in both a structured and unspecified procedures environment. Theirpurpose was to identify whether structured procedures, both with andwithout the aid of computers, led to better process or decisionperformance. They also examined the procedures the unspecified teamsused to determine what and how much of the current doctrinal processwould be considered. Part of their findings indicated that thoseofficers conducting tactical planning in an unspecified environmentfrequently failed to follow the doctrinal process:

Unspecified teams did not perform the task as 'chose teams whowere required to follow the procedural estimate guidance.Unspecified teams left out steps, did not perform steps in asobjective a manner as the structured teams, and repeated steps.16

Fallesen, Carter et al. (1992) concluded from their controlledexperiment that there were numerous problems that led each group todeviate from the doctrinal approach. Perhaps the most revealing ofthese concerns discovered during discussions with the participants wasthe general belief that the current doctrinal procedures were notuseful:

The application of a systematic approach is challenged by thecomplexity of the decision situations and procedures . . . and thepotential for lack of belief in the usefulness of the procedures.17

Lussier, Sollick and Keene (1992) developed a group planning andresource problem called VARWARS to evaluate the progress of C A S ~(Combined Armed Services Staff School) students at Fort Leavenworth anddetermine how effective course instructors were at teaching thedecision-making process. A startling revelation of their study was thatgraduates, during the progress of the course, had somehow developedworse analytical skills as prescribed by the current doctrinal process


30/156

than those students who were just beginning the course. They found thatperhaps one conclusion to explain this phenomenon was the studentsincreased focus on specific parts of the process instead of the qualityof the decision:

The major findings were that the graduates performedsignificantly worse than the entrants and that they did not use theproblem solving techniques which they had been taught. . . . [A]possible explanation for the lower scores of the more advancedgroups is that they are mistakenly emphasizing the wrong aspects ofthe problem.18

Halpin and Keene (1993) conducted a large and extensive surveyof combat leaders following DESERT STORM on a variety of topics, one ofwhich included the tactical planning process. Survey respondentsincluded over 1864 officers (52% captains, 21% majors), 478 NCOs(including 111 Sergeants Major) and 58 warrant officers. The majorityof the respondents were from division or lower echelons. Halpin andKeene reported that approximately 85% of those surveyed indicated thatthe current doctrinal procedures were adequate. However, there appearedto be a strong "minority viewpoint" that clearly defined the lack ofusefulness of the process once combat operations began.19 A sample ofthis viewpoint is provided by commanders at both company and battalionlevel

No process was used after combat operations were initiated.Missions were initiated with immediate response required. Pre-combat estimate process was adequate.

Company cdrZ0The estimate process worked well during planning, but the speedof offensive operations forced me to analyze the situation withlittle input from other staff members and make a decision.

Battalion cdr21


31/156

Grossman (1994) studied command and control of battalions andsmaller units at the NTC. Grossman analyzed the THPs (Take HomePackages) of units that trained at the NTC in 1989 (jusc prior to DESERTSTORM) and 1992 (the most recent), conducted in-field observations,administered surveys to participating units, and examined taperecordings of the task force command nets during preparation andexecution. His findings found that a majority of the units failbecause the task force staffs do not function well as a group or followeffective procedures to generate an adequate plan to defeat the OPFOR

(opposing force). Synthesizing information, tracking and reportingcritical information, and making good use of existing intelligence werealso highlighted:

Task Force staffs have problems generating adequate plans,managing battle preparations, and influencing the execution of thebattle. . . . Sixty-five percent of the plans generated by the TFcommander and staff are inadeq~ate.~~

summary. Observations at the CTCs, numerous independentstudies, and survey responses from combat leaders indicate many exampleswhere the current doctrinal process is not adhered to during practicalapplication. Again, numerous outside factors may influence the abilityof commanders and staff to stick to the stated doctrine; experience,level of training, team cohesion, and lack of time to name just a few.What does appear evident to researchers is that whatever the situation,whether actual combat, the CTCs or simulations, when the doctrinaldecision-making process is put to the test, it rarely survives intact.


32/156

Management of the ProcessIf empirical research suggests that commanders and staffs are

not following the process, perhaps there is one variable that is easilyobserved--the commander and staffs' efforts to manage the process.Current doctrine is not precise in describing how the commander managesand controls process. Doctrine places the commander's own estimateas a subset of the DDMP but provides little detail on how the commander"drives" the overall process. Either by default or because hecoordinates the efforts of the staff, this role of managing the processcurrently falls to the chief of staff/executive officer. If aiding thecommander to make a decision is the ultimate goal, then managing all thecritical activities to its logical conclusion may be as important as theprocess itself.

At least one major ARI study, as well as observations fromDesert Storm veterans, highlight the difficulty of managing the process.Thordsen, Galushkaet al. (1990) conducted a knowledge elicitation studyof military planning to primarily discover how often leaders makedecisions without concurrently comparing options (a process defined asRecognition-Primed Decision Making (RPD)), but also to examine anyobserved failures in planning behaviors. The controlled experiment wasconducted during a CPX at Fort Hood, Texas using observation, audiorecording, mapping, and coding of RPD behaviors. Their efforts todocument RPD behaviors were frustrated by the overriding observationthat management of the overall process was very weak and staff memberswere easily distracted:

The management of the process would appear to be as critical asthe actual planning itself. Military planners are specifically

24


33/156

trained to work with the domain specific knowledge but receiverelatively little training in the management of the process. . . .Our analysis showed 64 major transitions between objectives ofdiscussion during the five-hour planning session.23

Thordsen, Galushka et al. further observed that not only diddistractions unhinge the staff planning on numerous occasions, but alack of overall guidance and focus from the executive officers proved tobe just as debilitating:

The battle managers we observed simply jumped into the process,started generating and evaluat~ng ptions, and finished when theyran out of time. We dld not see any attempts to manage the use oftime. We did not observe anyone saying "we've got 5 hours here,and issues X, Y, and Z are the most important, so let's start withX but try to reach closure within 40 minutes."24

After observing this failure to manage the process Thordsen,Galushka et al. concluded that perhaps one of the biggest problems withthe current process is an inability by those executing the process tostay focused on the task at hand. Whether the process itself causesbattle staffs to become easily compartmented and distracted, or theobvious inability of executive officers to keep everyone focused on theend state, was difficult to determine. Thordsen, Galushka et al. foundthat whatever the reason, management of the process was a majorconsideration:

This emphasizes the need for the planners to fully understandexactly how the process unfolds, including all the distractions,pitfalls, and diversions. This is especially true if we want themto be able to manage this process as well as possible. Many of thedistractions were subtle and they occurred without any verbalizedawareness on the part of the planners, further handicapping theperson responsible for managing the planning session. 5

Management of the process also appears to have been a concernduring DESERT STORM as evidenced by a sampling of survey responses inHalpin and Keene's report (1993):


34/156

When time was short the S-3 limited the options for wargaming thusreducing time to wargarne.

Battalion ~ 0 ' ~We did what always is done but gets a "no go" at the NTC: thecommander and S-3, assisted by the S2 and FSO on the fringe,built a plan and execution matrix as quickly as we could.There was only one course of action wargamed based on the Cdrand S3 having to work without an accurate Intel picture and nofire plan from above. We established our scheme of maneuverbased on 2 days experience with the enemy and our knowledge ofour own strengths and weaknesses.

Battalion cdrZ7Summary. If commanders and staffs are not following the

procedures, then the obvious question is who should be responsible for

managing the process and producing the products that ultimately aid thecommander in making a decision. The current doctrine is vague and doesnot clearly explain how the coqander "drives" his own decision-makingprocess--he is only accountable for the end result. Although greaterdiscussion is provided in draft E l l 101-5, the chief of staff/executiveofficer remains largely responsible for managing the overall process.Minimizing distractors is a dynamic endemic to any team effort and doesnot mean that a given process is necessarily flawed. Yet, if trainingand experience play a big part in keeping the process focused, thenperhaps researchers should ask if the executive officer is the besttrained and most experienced individual to manage the process--or shouldit be the commander? Doctrine currently permits the commander to takeonly a nominal, if not entirely passive, approach to controlling "his"decision-making process.


35/156

Inflexible ProceduresFailure to follow the process or manage all its activities can

possibly be attributed to training and experience. Yet, when applied by

the best trained and most experienced combat leaders any process issuspect if it is not practical for every situation. In order to becritical of the process the researcher must first draw a cleardistinction between an inability to correctly apply the process orwhether the process itself is flawed. ARI research and survey commentsfrom battalion commanders in the field appear to support the suppositionthat the problems with the process are more than simple training andexperience.

Brezovic et al. (1990) conducted a study of Armored OfficerBasic students and instructors to identify the types of decisionstrategies used by individuals with varying levels of experience. 21students and 9 instructors were examined during force-on-force trainingexercises. Subsequent interviews helped the researchers isolate 57decision points to determine how each individual deliberated to make adecision. Brezovic et al. concluded that the prescriptive model usedat higher echelons was not applicable at the tactical level because itdid not allow for flexibility in time critical situations or address thenecessity to make decisions when very little information is available:

One clear barrier to the uncritical adoption of a formal model tothe tactical environment is the emphasis on identifying as manymajor problem dimensions as possible BEFORE executing any action.If this analysis were followed as typically described, the resultcould well be an untimely termination of cormnand long before aresponse could be generated. 8

Fallesen, Carter et al. (1992) in the same controlled experimentdiscussed earlier also concluded that the current doctrinal procedures

27


36/156

were inflex~ble. Their assessment was so strong as to infer that combatleaders under stress and running short of time do not use the currentprocess--a serious indictment on any doctrine if it only applies intraining. Any gains realized by the careful consideration of optionsand weighing of all available information before making a decision arequickly overshadowed by the complexity of the environment and a shortageof time:

It is likely that experts would not (and do not1 tolerate aslow, rigid, and arduous process which might only showinconsequential differences in combat outcomes or with outcomesthat are largely uncertain. . . . Using a decision analyticapproach, as complicated as a weighted, multi-attribute utilitymatrix or as simple as summary columns of pluses and minuses, canbe misleading for complex, dynamic tactical problems. 9

Lussier and Litavec (1992) surveyed 48 battalion and squadroncommanders returning from NTC, JRTC and/or DESERT STORM, 25 of whichwere graduates of TCDC (Tactical Commanders Development Course). Theirprimary goal was to assess the relative adequacy of the TCDC course, butclosely tied to this evaluation was an assessment of the Army'sdeliberate decision-making process. Numerous respondents stated thatthe process was flawed and not applicable under combat or simulatedcombat conditions because it was too rigid, time consuming, and failedto show how or when to abbreviate steps:

If planning is being done during execution in a changingenvironment, the decision-making process must be abbreviated evenfurther. This is a problem, commanders admit, with which theygenerally have not coped well. . . . Doctrinal publications teach astaff estimate process involving comparison of several courses ofaction. Add in briefbacks, wargaming, rehearsals, executionmatrix, etc. and there is too much for the commander and staff todo. A11 these techniques need to be carefully considered.Doctrine should provide a standard timeline describing what must bedone and what can be dropped, including some guidelines on how muchtime should be spent on each product or activity.30


37/156

Summary. Researchers and comments from the fie1d.suggest thatthe current prescriptive decision model as described by doctrine islacking. Fallesen (1993) summarizes that the traditional viewpoint isthat the process is acceptable, but needs to be performed faster andmore efficiently.31 Admittedly, it is difficult for military theoriststo completely dismiss an analytic approach to solving complex tacticalproblems. Afterall, this approach has evolved over 50 years and isfirmly entrenched in the mindset of Army leaders. Yet, researchsuggests that perhaps change is required. In any basic problem solving

format an individual or team must still gather the facts, determine theproblem, consider the variables that affect the outcome, and then make adecision. This is the essence of any decision-making process. Perhapsthe issue then becomes what part of the process is the most important,and what can it be streamlined without removing any basic elements thatlogic demands. Better training and more cohesion may fix some of theproblem, but this stili does not address the doctrinal problem of alockstep methodology unsuited for time-constrained environments.Success will demand greater flexibility and doctrine does not address insufficient detail how to effectively abbreviate the process.

Excessive Time DemandsClosely tied to observations that our current doctrinal

procedures are inflexible is the major impact that time has in everytactical situation. Often, time is attributed throughout history as thesingle most decisive factor that determined victory or defeat.Naturally, in any violent contest between opposing forces, with all


38/156

things fairly equal, that side which makes "quality" decisions quickerwill have a distinct advantage. A common tenet espoused throughout the'Army is "A poor plan executed on time is always better than a perfectplan executed too late!" Certainly this lesson of history did notevolve without impetus. Perhaps experience has taught commanders thataction counts more than making the "best" decision, and if that is true,then time is definitely an ally.

Thordsen, Klein, Michel, and Sullivan (1991) examined a class of62 students in Advanced Warfighting, at the U.S. Army's Command andGeneral Staff College (CGSC) The researchers observed student planningcells during three separate planning exercises and specifically focusedon ten distinct categories using knowledge elicitation methods to assessdecision-making behavior. The primary goal was to determine if theirknowledge elicitation methods could be used in future training exercisesto provide commanders and staffs after-action feedback of planningscenarios. During the course they also reached numerous conclusions onstudent behavior during application of the current doctrinal decision-making procedures. Students had difficulty making decisions in a timelymanner. The researchers found that this was due to the students impulseto seek the "perfect" plan rather than make a timely decision notnecessarily supported by all the available information:

It is very easy to be overtaken by events when the planning cell isnot looking far enough ahead. This can be tricky, because thenatural inclination is to wait until you have all the informationnecessary to make a relatively risk free plan.32

Fallesen, Lussier, and Michel (1992) documented known empiricalresearch on the command and control process to provide lessons learned


39/156

for CGSC instructors to improve the process as it is taught in theclassroom. They reviewed numerous sources, from simulations to fieldobservations, and concluded that much of the planning time is wasted bybattle staffs during the attempt to generate and evaluate options. Whattheir research discovered was that most battle staffs do not developseparate and distinct courses of action for comparison and evaluation,but instead place most of their emphasis on what they "believe" to bethe best option and give little credibility to the other options. Theother options are only developed because current doctrinal procedures

require it and time constraints did not allow for a equal considerationto each option:

The traditional estimate process addresses the preparation ofmultiple courses of action (given enough time), analyzing eachindependently, comparing the results, and only then selecting anoption. But today's battlefield will rarely provide enough time tofully develop multiple courses of action. Planning and preparationtime can easily be wasted by spending too much time on "throw-away"options, at the risk of acting too late and losing theinitiative. 3

Lussier and Litavec's (1992) survey of battalion and squadroncommanders found that time was the overriding factor to most, if not allthe commanders when applying the deliberate decision-making process.Across the board, their responses indicated a general dissatisfactionwith the current procedures and their inability to abbreviate theprocess:

Almost all commanders were adamant that the Army's decision-making processs was too ponderous for use in limited timesituations, and to attempt to use this process in those situationswas a mistake. The process was workable and beneficial in theDesert Shield preparation phase; however one Desert Storm began, itwas not.34


40/156

Commanders distinguish two situations: limited time situations,with only a few hours of planning time available, and executionsituations, where mission planning is occurring at the same time asexecution. In the latter case, the changing tactical environmentmakes the doctrinal decision-making process even less applicable.Commanders know they must truncate the process. Most commandersbelieve they are not given much doctrinal help in doing thattruncation. 5

Halpin and Keene's (1993) survey of the combat leaders returningfrom DESERT STORM found similar responses concerning the practicalapplication of the current doctrine in time constrained environments:

There were times when the situation just didn't allow for bigplanning sessions--leaders should receive more training for this in[the] advance courses!

Battalion F S O ~ ~As the battlefield became more fluid, short notice FRAGOs becamethe norm. Execution details were worked out as the missiondeveloped. Overall, the staffs responded well to the flexiblesituation.

Assistant ~ - 3 ~ ~Corps plans were not timely, forcing Division to jump through ahoop in order to execute; 1/3 - 2/3 rule became 7/8 - 1/8.

LTC, L N O ~ ~Too fast paced operations. The 1/3 - 2/3 rule for planning was notapplied at Bde level. This impacted on planning/prep time at Bn &Btry level.

Battalion ~ d r ~ ~Smary. The overwhelming conclusion from researchers and

combat leaders is that time is critical in the planning process, and thecurrent process is too ponderous to apply in complex and stressfultactical situations. If for no other reason than to prevent combatleaders from arbitrarily dismissing the entire process out offrustration, the exact opposite effect desired of any decision model,doctrine must prescribe how to effectively abbreviate the process.

32


41/156

Formulation of AlternativesConventional decision-making theory focuses on selecting the

"best" option that provides the maximum result with the minimum cost.Selecting that "best" option is routinely discussed relevant toavailable information and not in respect to time constraints. However,recent theorists have started to explore a concept, called "NaturalisticDecision Making", that focuses on selecting a "feasible" option thatsatisfies the problem based on time available, information queues, andthe expertise of the decision maker. The assertion is that decisionmakers, operating in cognitive domains that are familiar, will recognizecritical elements of information based on experience, which leads themin turn to select a feasible option to fit the situation. The key isexperience. The more experienced the decision maker, the better thedecision.

Recent research indicates that in tactical environments thedecisions made using a "Naturalistic" approach have no less impact onthe outcome than the traditional theory of generating and selecting the"best" option. In either case, within the tactical enviromnt, thecommander who is making the decision is still influenced by recognizableinformation queues. If the assumption is made that the commander is themost experienced decision maker, then the "committee approach" by thestaff to generate viable options will not significantly influence whatthe commander, through his experience, recognizes as his first and mostfeasible option. Waiting for the staff to generate those optionsbefore the cormander makes a decision, however, is time consuming.


42/156

Naturalistic decision theory advocates skipping this step. In atactical situation the commander's "gut" call, based on his experienceand available information, will still probably be the "best" option--butthe decision is arrived at much faster.

Klein and Calderwood (1990) reviewed three years of researchconcerned with how experienced personnel make decisions in operationalsettings characterized by real-time information processing, shiftinggoals, and high-risk consequences. The actual study domains selectedwere determined to correlate to general military decision-making, these

included urban fire ground commanders, wildland fire incidentcomanders, and U.S. Army tank platoon leaders. A large part of theirreview presents current theoretical thinking which now opposes earlierviews on serial generation:

The work of Simon (1955) is probably the best known discussionof a serial generation and evaluation strategy in the behavioralscience literature. Simon described the use of satisficing as ameans of quickly and efficiently finding an effective option.Satisficing is a process by which choices are evaluated one at atime until a satisfactory one is found, and then implemented. Itdiffers from an optimization strategy in which virtually alloptions have to be generated and evaluated in order to determinewhich is best. 40

Klein and Calderwood (1990) thus concluded:Should Proficient Decision Makers Generate as Many Options as

Possible? From the perspective of recognitional decision-making,the answer is "No."

This recommendation is heard from both decision researchers(e.g., Gettys, 1983) and practitioners writing popular books andarticles (e.g., Janis & Mann, 1977). In the time-pressuredenvironments we studied, there simply was not enough time to followsuch advice. It takes time to generate and evaluate options, anddelays may be intolerable. In addition, the situation may shiftduring the analyses so that the whole process has to start overagain. Even in the absence of time pressure we rarely observedproficient decision makers trying to generate large sets ofoptions. 41


43/156

Geva ( 1 9 8 8 ) studied three combat decision situations during thesecond and third days of the Yom Kippur War in the Suez Canal region.He interviewed commandezs at every echelon from division to battalionand reviewed numerous other printed sources in an effort to assess howdecisions were made under the stress of combat. With only oneexception, Geva determined that those commanders examined did notgenerate and concurrently compare options before making a decision asdoctrine allowed. The Israeli commanders usually selected the firstcourse of action generated. Geva asserted that in most'cases experienceled the commanders to believe that the first course of action developedwould ultimately be the best decision:

In three cases there were indications of more that one alternativecourse of action. Only in the second case the alternatives werecompared prior to casting the decision. In all other cases, thefirst raised alternative was adopted as the decision. Otheralternatives emerged subsequently, either when the originalalternative was reported or perceived ~nfeasible.~~

As previously mentioned, Brezovic et alls.,'s (1990) researchinvolving the armor platoon leaders in the basic course at Fort Knoxappears to support Geva's observations:

We have found a preponderance of cases where the decision makerrelies on situational awareness, that is, the ability toimmediately ascertain the important features of a decisionsituation and to derive the appropriate implication^.^^

Brezovic et al. '(1990) reviewed the reaction of these youngarmor platoon leaders making decisions in combat simulated environmentsand concluded:

The conclusion one draws from these few accounts of militarycommand in action is that there is little place for the carefulreasoned approach, insensitive to time, that has been proposed asthe model for human decision-making. On the contrary, aninappropriate emphasis on the generation of options over the search


44/156

for patterns in the decision environmedt seems a clear prescriptionfor disaster. 4

Also previously discussed in this review, Thordsen, Galushka etal. (1990) in their knowledge elicitation study of military planningobserved a very distinct difference between what U.S. Army doctrineadvocates and what is believed by some to actually occur in practicalsituations:

We have shown (Klein, Calderwood, & Clinton-Cirocco, 1986;Klein, in press) that "experts" in natural decision-making settingsoperating under various levels of time pressure and stress do notfollow the classical decision process. These individuals have beenshown to approach the decision-making process by addressing onesingle option at a time. Often this involves making an automaticdecision, with little if any deliberation. When multiple optiondeliberation does occur, it tends to be along serial, rather thatconcurrent, lines. 5

Thordsen, Galushka et al. (1990) further concluded in'theirstudy that decisions made by their control groups were performed withoutgenerating and concurrently evaluating options as U.S. Army doctrinesuggests:

We found 96 percent of the identified decisions (26 out of 27)were RPDs where there was no concurrent deliberation. That is,multi-attribute weighting of options, decision analysis and/orBayesian strategies did not come into play.46

Lussier, Solick, and Keene (1992) observations, using theVARWARS model also previously discussed, reached very similarconclusions:

Neither entrants nor graduates followed the method taught, asonly one group proposed more than one adequate alternative, andthat group failed to complete the process by analyzing bothalternatives. The primary process of problem solving employed byboth types of groups was to develop a single solution, modifying itas necessary and rejecting it only if it became completelyunworkable. 7


45/156

Fallesen, Lussier et al. (1992) in their collaborative effort todiscuss the tactical command and control process strongly felt thatarriving at an early decision, without consideration of options,actually offered practical advantages that outweighed the negligiblebenefit of comparing similar courses of action:

When is one course of action enough? Each tactical situationis unique. In some cases, the number of feasible options islimited, or a single "best" option may be apparent to theexperienced decisionmaker. In other cases, several distinctoptions may be feasible, but in the cormnander's judgment they areso close in value that it is not worth the time and effort requiredto determine which is "best." In any case, time constraints maymake the analysis of several options impractical. In theseinstances, it is best to quickly arrive at a preferred course ofaction. This gives maximum time to the staff and to-subordinatecommands for the planning, coordination, and rehearsal that isessential for the accomplishment of any course of action.48

Summary. Because a significant criticism of the current processis the amount of time required to complete all the steps, a large partof the current research advocates a reduced emphasis, if not thecomplete elimination of, the concurrent evaluation of separate coursesof action prior to the commander's decision. Some evidence indicatesthat on those occasions when staffs have considered separate options,those same options are generally so similar that little benefit wasgained by the effort. As a rule, observations on the process tend tosupport the trend that the first option considered and developed isgenerally the "best" option anyway. Most of the time staffs developedother options only to satisfy the requirements of the process and neverseriously considered the remaining options as a potential solution. RPDtheory would also seem to support this observation. Experience, more sothan the careful consideration of numerous options, appears to be the


46/156

real measurement of the quality of a decision. The more experienced thecommander and his staff, the better the decision. Concurrent evaluationof separate courses of action using a weighted, multi-attribute utilityanalysis does not necessarily improve the quality of the decision.

Wargaming and SynchronizationCurrent research also supports the observation that wargaming

and synchronization are not performed well. If no plan survives intactafter combat begins, then prudent commanders and staffs should considerthe impact of possible branches and sequels to the original course ofaction. Researchers and commanders in the field observe that despitethe importance of this step this is rarely achieved to satisfaction.Several reasons are presented to explain this deficiency which include alack of experience, too little time, too much focus on picking the bestoption, and the staff becoming too enamored with the synchronizationmatrix.

Fallesen and Michel (1991) observed a recent BCTP warfighterexercise to provide observations on the practical application of thedeliberate planning process at the corps, division, and brigade level.The subjects of the research were CGSC students and not actual unitleadership, so the test group actually negated the benefits normallygained by experienced decision makers executing the process. The

researchers were able to observe a large test group recently schooled inthe process, but who generally lacked the experience and would be lessinclined to deviate from any formal steps in the process. As such, theresearchers gained considerable benefit by observing what they felt


47/156

would be a more rigid application of the process. Fallesen and Michelobserved:

The students gave very little thought to how the initial planscould go wrong. They were never observed considering branches ineither enemy actions or their own actions. . . . There was verylittle detailed war gaming. War gaming did not involvequantitative estimates of relative combat power, identification ofcritical events,. and projection ,of engagement results.49

Fallesen and Michel (1991) also observed how students becameenamored with the synchron~zation atrix as essential to the process,yet failed to cons~der he primary reason why the matrix was used--torecord the results of war gaming and serve as a basis for comparison ofagainst other courses of action. The mental focus was for each planninggroup to simply complete the matrix and not how the matrices, ifdesigned with similar parameters, could help the commander' see thedifferences between the two options:

At Corps, two teams were used each to complete asynchronization matrix on one concept with the purpose, in part, tofurther consider which was the better COA. The two teams did notcoordinate in advance how each one would do the matrix, so whileone team was doing synchronization at (projections of battle eventsfor) 4 hour intervals, the other was doing it at 12 hour intervals.One started at H hour, the other started about H-24 hours. Wherethe two efforts ended up for comparison was not observed, but themismatches of different bases of comparison is evident. 0

Thordsen, Klein et al. (1991) also observed the inherentproblems with using the synchronization matrix as a solution forconsidering branches and sequels of a particular course of action andcomparing it to other courses of action:

As good as the matrices are, we do urge caution in that theyhave the potential to lull one into a false sense of security dueto the great amount of detail they require to be successfullycompleted. This amount of detail can misguide one into believingyou have covered everything, when in fact, there will always besome things that are missing. Another point of caution is that


48/156

this same amount of detail can potentially result in an informationoverload condition where there is so much detail that criticalmaterial gets burled in the volume.51

Lussier and Litavec's (1992) collection of commander's comments,

in support of earlier discussion of concurrent evaluation of severaloptions, advocates that the "optimal" decision is not as important as a"workable" decision that is arrived at earlier, better wargamed,synchronized, and executed. The inference here is that less time spenton picking the best solution can be better spent on quality wargaming ofthe actual decision and a better appreciation of all the possiblebranches and sequels:

They emphasized that the important thing is how well plannedand executed the mission is, not which course of action is chosen.A suboptimal plan . . . that is well executed is much better than aperfect plan that arrives to late. Increased decision-making timedirectly reduces planning time, rehearsal time and subordinateplanning and preparation time. These latter activities have muchhigher payoff than the possibility of arriving at an incrementallybetter course of action.52

Fallesen, Carter et al. (1992) not only observed that wargamingand synchronization were the more important steps of the process, butalso the most difficult and required the most experience to complete.This condition was further exacerbated by a general lack of acceptedtechniques for using automation and simulation to observe theapplication of combat power in relation to time, space, and expectedoutcomes

Wargaming and comparison were the most difficult steps in theexperimental task. Participants generally lacked the knowledge andexperience to make battle projections. They reported that neitherdoctrine or instruction provides adequate guidance for making wargaming p~ojections.3


49/156

Summary. Current procedures as outlined in ST 101-5, clearlydefine the purpose of war gaming and the development of thesynchronization matrixas the last step prior to comparing the separatecourses of action. The most recent research and observations from thefield indicate that the primary purpose of war gaming should be therefinement of the intended course of action and not as a process forselecting between recommended options. War gaming is in essence anappreciation of intended outcomes as they relate to time and space--action, reaction, and counteraction. There is some evidence to suggestthat this understanding of how a course of action may unfold is aproduct of experience. Current procedures imply but do not dictatethat the comander participate in the war gaming session for everyrecornended course of action. This is certainly time consuming and hemay simply review the results of the synchronization matrices developedby the staff. The synchronization matrix is only a planning tool torecord results. When the commander does not participate in the wargaming process, his experience precludes the best possible visualizationof the expected outcomes.

Some researchers and commanders feel the best focus for wargaming, with the commander's involvement, is the intended course ofaction. If a decision can be made quickly, greater benefit is gained bythen concentrating on all possible branches and sequels of the selectedcourse of action. The evidence seems to suggest that staffs simply donot war game adequately. Developing a synch matrix for each recommendedcourse of action is a time consuming process. Current observationsindicate that staffs may become too enamored with completing the


50/156

synchronization matrix, and fail to appreciate t he importance of the wargaming process for visualizing the intended outcome. Losing sight ofthe intended purpose of war gaming is further compounded by th e lack ofinvolvement by th e commander.


51/156

Endnotesl ~ o n . Faklesen, Overview of Army Tactical Planing Performance

Research (Alexandria, VA: U.S. Army Research Institute for theBehavioral and Social Sciences, 1990), 1-19.

'u. 3. Army, FM 101-5 (draft), command and Control for Commandersand Staff (Washington, DC: Department of the Army, 1995).

"u. S. Army, ST 100-9, The Command Estimate Process (FortLeavenworth, KS: Army Command and General Staff College, 1986).

15~ex . Michel and Sharon L. Riedel, Effects of Expertise andCognitive Style on Information Use in Tactical Decision Making(Arlington, VA: U.S. Army Research Institute for the Behavioral andSocial Sciences, 1988), 26.

16~on . Fallesen, Charles F. Carter, et al., The Effects ofProcedural Structure and Computer Support Upon Selecting a TacticalCourse of Action (klington, VA: U.S. Army Research Institute for theBehavioral and Social Sciences, 19921, viii.

18~ames . Lussier, Robert E. Solick, and S. Delane Keene,Experimental Assessment of Problem Solving at the Combined Arms andServices Staff School (Arlington, VA: U.S. Army Research Institute forthe Behavioral and Social Sciences, 1992), 15-16.


52/156

19stanley M. Halpin and S. Delane Keene, Desert StormChallenges: An Overview of Desert Storm Survey Responses (Arlington, VA:U.S. Army Research Institute for the Behavioral and Social Sciences,-.l993), 18.

22~on rossman, Battalion-Level Command and Control at theNational Training Center (Santa Monica, CA: RAND, Arroyo Center, 1994),21.

23~arvin hordsen, Joseph Galushka, et al., A KnowledgeElicitation Study of Military Planning (Arlington, VA: U.S. ArmyResearch Institute for the Behavioral and Social Sciences, 1990), 39.

26~alpin nd Keene, 15.

28~hristopher . Brezovic, Gary A. Klein, and Martin Thordsen,Decision Making in Armored Platoon Command (Arlington, VA: U.S. ArmyResearch Institute for the Behavioral and Social Sciences, 1990), 4.

29~allesen, arter, et al., 65-90.30~ames . Lussier and Douglas J. Litavec, Battalion Commanders'

Survey: Tactical Commanders Development Course Feedback (Arlington, VA:U.S. Army Research Institute for the Behavioral and Social Sciences,1992), 18-20.

32~arvin . Thordsen, Gary A. Klein, et al., Methods forProviding Direct Feedback About Decision Processes for Command andControl Classroom Exercises (Arlington, VA: U.S. Army ResearchInstitute for the eehavioral and Social Sciences, 1991), 18.

33~on allesen, James W. Lussier, and Rex R. Michel, TacticalCommand and Control Process (Arlington, VA: U.S. Army ResearchInstitute for the Behavioral and Social Sciences, 19921, 8.

34~ussier nd Litavec, 16.

36~alpin nd Keene, 17.


53/156

40~ary . Klein and Roberta Calderwood, Investiqations ofNaturalistic Decision Making and the Recoqnition-Primed Decision Model(Arlington, VA: U.S. Army Research Institute for the Behavioral andSocial Sciences, 1990), 13.

42~. eva, Executive Decision Making Process During Combat: APretest of the Content Analysis Schema (Zikhron Ya'akovi, Israel:Israeli Institute for Military Studies, 1988), 32.

43~rezovic t al., 1-2.44~bid., .45~hordsen, alushka, et al., 41.46~bid., 42.47~ussier, olick and Keene, 10.48~allesen, ussier, et al., 8.49~on allesen and Rex R, Michel, Observations on Command and

Staff Performance During CGSC Warrior '91, 1991), 5-9.

51~hordsen, lein, et al., 17.52~ussier nd Litavec, 17.53~allesen, arter, et al., 89.


54/156

CHAPTER THREEA PROPOSED COMBAT DECISION-MAKING PROCESS

The review of literature demonstrates that commanders and staffsexperience difficulty when using the deliberate decision-making processin time-constrained environments. Evidence indicates that the currentdeliberate process may be too inflexible to compensate when planningtime is limited. Observations and interviews from commanders in thefield also indicate a majority viewpoint that doctrine does notadequately address how to effectively abbreviate the process.

Draft FM 101-5 describes only one decision-making process--thedeliberate decision-making process. Draft FM 101-5 further stipulates,"It is extremely important that the deliberate process be usedinitially, returned to whenever operations tempo allows, and practicedduring all training events."l It is widely accepted by militaryprofessionals that individuals and units fight as they are trained. If,as draft FM 101-5 indicates, commanders and staffs should train solelyon the deliberate decision-making process, then our doctrine may beinadequately preparing commanders and staffs for the actual conditionsof combat.

Given the documented observations, it is entirely appropriate toexplore the question of "when" and "how" to abbreviate the deliberatedecision-making process, or provide a more suitable decision-makingmodel that better prepares commanders for combat. The purpose of this

46


55/156

chapter is to present an alternative to the DDMP that the researcherbelieves more adequately addresses decision-making in time-constrainedenvironments. If subsequent analysis or further research validate theproposed combat decision-making model, the researcher believes thismodel should become the Army's focus for training. If individuals andunits do indeed "fight as they are trained", then the proposed model mayserve as a better framework for making decisions in combat.

PROPOSED COMBAT DECISION MAKING PROCESS

Fig. 2. Annotated Steps of the Proposed Combat Decison-Making Model.

47


56/156

As currently written, the deliberate decision-making processconsists of nine annotated ste&,>as outlined in figure Theproposed combat decision-making process shown in figure 2 consists ofeight annotated steps. Each is explained and discussed further as theydiffer from the current DDMP prescribed in draft l?M 101-5. Both theproposed combat model and the DDMP are then compared in chapter fourusing the criteria discussed in draft FM 101-5 that defines an effectivedecision-making process.

Mission Analysis (STEP 1, STEP 2, STEP 3 )With minor exceptions, the first three formal steps in both the

deliberate and the proposed combat decision-making process remain thesame. As is the case in any analytical process the first steps alwaysconsist of gathering all known or relevant information, makingassumptions, and recognizing or defining the problem. This sequence oflogic remains consistent in both the deliberate and the proposed combatmodels. Despite any assistance from the staff, the commander bears theultimate responsibility for defining and communicating the end state.Whether in a peacetime planning environment relatively unconstrainkd bytime, or under the stressful conditions of combat, the commander mustclearly "see" what must be accomplished. No single process or decisionmodel, despite its effectiveness, can rescue a commander from hisfailure to correctly identify the problem or clearly define the misstonwith his intent and envisioned end state.

The proposed draft FM 101-5 states that within the framework ofthe deliberate decision-making process the commander and each staff


57/156

member prepares an e~timate.~ he synthesis of information provided bythese estimates initially comes together in mission analysis. Althoughdraft EM 101-5 further states that the commander "drives the process",the doctrine purports that the commander and staff are executing twoseparate processes designed to mutually support the other.4 Thisseparation in the process is illustrated below in figure 3.

DELIBERATE DECISION MAKING PROCEDURETask R.S.md

ImmMrItD Colnmnd.r m san+ProP0.M

II

Fig. 3. The Deliberate Decision-Making Procedure extracted from ST 101-5, Command and Staff Decision Processes, pg 1-2-6.

The proposed combat model differs with this separation of thecommander and staff during mission analysis. In the combat model, thecommander and staff perform mission analysis together--not separately.Essentially, there is no parallel process or separation between the

4 9


58/156

commander and his staff throughout the entire combat model. Not onlydoes the commander "drive the process", he actively leads the staffthrough each step. This is a distinctly different approach from thecurrent DDMP.

Additionally, the proposed combat model places greater emphasisthat mission analysis is both continuous and cyclic. In DDMP, missionanalysis is essentially performed only once during the process. Theproposed combat model purports that mission analysis should beconstantly reassessed duri

combat decision making process at the tactical level

Documents